Bayesian Embeddings for Few-Shot Open World Recognition

TL;DR

This paper introduces FLOWR, a Bayesian non-parametric framework for open-world few-shot recognition that improves classification accuracy and novel class detection over prior methods.

Contribution

It extends embedding-based few-shot learning to open-world scenarios using Bayesian non-parametrics and demonstrates superior performance on benchmark datasets.

Findings

Up to 12% improvement in H-measure for novel class detection

Strong classification accuracy on open-world benchmarks

Effective combination of Bayesian priors with embedding pre-training

Abstract

As autonomous decision-making agents move from narrow operating environments to unstructured worlds, learning systems must move from a closed-world formulation to an open-world and few-shot setting in which agents continuously learn new classes from small amounts of information. This stands in stark contrast to modern machine learning systems that are typically designed with a known set of classes and a large number of examples for each class. In this work we extend embedding-based few-shot learning algorithms to the open-world recognition setting. We combine Bayesian non-parametric class priors with an embedding-based pre-training scheme to yield a highly flexible framework which we refer to as few-shot learning for open world recognition (FLOWR). We benchmark our framework on open-world extensions of the common MiniImageNet and TieredImageNet few-shot learning datasets. Our results show, compared to prior methods, strong classification accuracy performance and up to a 12% improvement in H-measure (a measure of novel class detection) from our non-parametric open-world few-shot learning scheme.